I'll describe VM's strengths throughout this paper, but will summarize them here: VM is the only operating system that combines the industrial strength aspects of robustness and scalability that one associates with mainframes, with the flexibility, low-cost, and accessibility associated with PCs. It has a proven track record for production computing, end-user interactive computing, and user enablement. Against a history of attack, it has survived only because it is so good at being fast, flexible, and efficient. What VM has not succeeded at is being recognised as a mainstream, strategic system.
The VM community has often encouraged VM in spite of IBM rather than with IBM's help. I exclude, of course, the VM development organization within IBM, and many other IBMers, who have valiantly supported VM.
VM's success over the years has been almost totally due to its fans (often called VM bigots), both within and outside IBM, who insisted on VM instead of IBM's officially sanctioned answers. The only comparable experience is Unix, which also rose to fame due to grass-roots enthusiasm. There must be something of enduring value in an operating system environment that inspires such loyalty.
Throughout VM's history, there have been IBMers telling customers that VM would shortly be phased out and replaced by a "strategic" IBM product. If IBMers tell you VM is ready for the junk heap, remind them how often that claim has been made, how foolish IBM has looked, and how customers moving off VM have gone to other vendors. Then, please pass their names to somebody in the VM organization so they can get an attitude readjustment from somebody responsible (I'll be glad to do it, but it's better when it's not hearsay).
This brings up one of my complaints about IBM: no other company has such a habit of disparaging its own product line. I'm certain that a derogatory public remark about the Private Client part of my company made by a member of the Capital Markets side would be treated severely. It amazes me that IBM tolerates public behavior - in front of customers - in which its staff makes insulting references to parts of its product line. This is unprofessional but widely practiced IBM behavior. It is inconsistent with a "world class company". It certainly is not an inducement to buy IBM products. IBM is in many ways a fine company, one of the world leaders, but this particular behavior is one that I find inexplicable.
If you have a long memory, you'll recall a list of inadequate products that IBM wanted customers to use instead of VM. IBM designated VSPC (Virtual Storage Personal Computing) on OS/VS as the "strategic interactive system". This lasted until MVS's TSO emerged as IBM's strategic interactive system, even though TSO was just barely interactive. A few years ago, IBM thought AIX/ESA would replace VM, and diverted talented developers from VM to work on a quixotic attempt to port a complete Unix operating system (unlike the OpenEdition environments that run under MVS and VM) to mainframes and get customers to pay for it (the first part of that happened; not the latter). Subsequently, IBM's AIX (on RS/6000) and OS/2 systems were supposed to make VM unnecessary. Billions of dollars were spent on OS/2, which is almost a total failure. Now, talented VM developers and considerable resources have been diverted to Lotus Notes and Domino. I expect this will not be particularly successful either.
Simply stated, IBM has frequently been dead wrong about what technologies were worthy of being deemed strategic. This has soured me on so-called "strategic planning"; a plan or product is designated "strategic" if you can't think of a way to justify it based on its merits. At your own office, ask proponents of "strategic" initiatives to justify them on business return, not on the basis if a slogan.
IBM's current system-to-replace-VM is Lotus Notes, and IBM has embarked on a attempt to move from OfficeVision/VM to Notes. I believe this will fail, at least in the sense of providing improved computing and communications for its users. IBM is also mistaken in thinking that VM and OfficeVision are synonymous.
If IBM really thinks that Lotus Notes is that good, then it should neither mandate nor subsidize its adoption within IBM. Let it compete on its merits. If IBM offices think that Notes provides improved service and better costs, then they'll adopt Notes without a top-down mandated answer. If they don't believe this, then perhaps there is something wrong with the idea. I think it's even less likely that Lotus Notes will be adopted by paying customers in the numbers that IBM needs.
Current estimates leaking out of IBM are that IBM has spent $9 billion on Lotus Notes and Domino (I'm writing this 1Q98); $3.5 billion for the acquisition and the remainder for development and marketing. It's been reported that Notes costs "only 50% more" than OfficeVision - providing you exclude the costs for upgrading the network and every PC! The total cost of ownership of Lotus Notes is clearly going to be far higher.
To those that say movement from VM to Notes is good business for IBM, I reply that it is unlikely that IBM will acquire the market share, or hardware and operating system revenue for the platform Notes runs on: IBM is obviously not the leader in personal computing. We buy the operating system from Microsoft, and the hardware is mostly from PC manufacturers like Compaq, Dell, or Gateway. We don't even call these machines "IBM PCs" anymore. The idea that we will see a large migration to OS/390-based Lotus products (where there would be a revenue stream to IBM for operating system and hardware) is laughable.
IBM has to understand that the industry leaders in this space are Microsoft's Office suite, notably Exchange, and general open-systems communications based on the Web, SMTP, POP, and IMAP. Lotus Notes is a poor third-place in this market. IBM bought an expensive proprietary groupware product at peak price, just as the industry dominating software vendor (Microsoft) and the Web provided highly attractive alternatives. I don't expect its revenues to ever justify its $3 billion investment or the subsequent $6 billion. In my opinion, most of this money has been wasted.
I think some, but too few, managers in IBM understand VM's value and unique capabilities. IBM is doing valuable development to support VM and increase its capabilities. Nonetheless, we should never relinquish to IBM all responsibility for VM leadership. Too much is at stake, and frankly, IBM has been wrong too often before.
IBM still finds it difficult to break its bad habit of promoting new technology for MVS, when the technology is equally applicable and frequently a better match with VM. The (initially) MVS-only announcements for the CMOS and Parallel/390 systems is an example, despite VM's far superior horizontal scalability. While the struggle for adequate device and processor support has been largely won, we still are forced to fight with IBM, for example, to provide badly needed improvements to VM TCP/IP (now well under way due to the VM development group) and OfficeVision/VM, and to get them to finally (and aggressively) market the P/390.
The computing industry follows only one vendor at a time with docile, sheep-like suspension of critical judgement. Twenty years ago that was IBM, but now it's obviously Microsoft. It's not likely that IBM will convince us all to abandon VM, yet stay on their product line instead of the products from the industry leading vendor.
IBM is not the industry dominating vendor of choice in the PC and workstation worlds. Their market share for selling PC servers has been on a declining curve, and their PC operating system, OS/2, is almost completely irrelevant and essentially abandoned.
It is foolish for IBM to assume we'll continue to buy from them if we abandon their existing products. My experience has been that we've spent our money elsewhere. IBM must therefore not abandon its traditional product lines. To do so would discard their customer value and excellent technology, and would speed the defection of their current customer base to non-IBM platforms.
Tomorrow's systems must compete on their merits in order to survive. VM, having survived on its competitive merits for all these years, is one of the best equipped of IBM's systems to do so. VM is also the only system IBM has that can compete both at the desktop and in the datacenter, providing IBM its sole unique advantage of unparalleled scalability and compatibility.
The biggest problem for VM isn't what VM can or cannot do, but the lack of marketing support from IBM. There's a complete lack of marketing VM products out in the field. There is almost no sign of branch involvement in the New York area for marketing VM products and solutions, and acts of actual cowardice in which a lucrative and visible market segment was abandoned to Sun. Part of the problem is that the branches don't understand the product line. Many only know MVS, and don't sell inexpensive VM solutions because it doesn't provide interesting commissions. Development groups within IBM have also abandoned VM, refusing to provide products comparable with MVS versions (such as a C++) or providing versions of their products on competitor's operating systems before doing VM. This is hard to understand.
As a side note, I must comment that IBM must have some strange programming practices if they are writing new S/390 software in such a non-portable way that they consider it a massive expense to port from MVS to VM/CMS or VSE. It is short-sighted thinking to produce new S/390 code that runs on only one S/390 operating system. If IBM wanted to, they could orient their software manufacturing to produce portable code, at least within the S/390 world.
IBM extended MVS-like marketing muscle to AIX and OS/2, both of which compete against far more successful and influential systems like Sun's Solaris and Microsoft's Windows. Despite having good technology (an operating system that comes with Rexx as a script language can't be too bad!) OS/2 became as irrelevant as any operating system can be, yet was showered with development and marketing funds until Gerstner finally shut it down. How ironic that OS/2 received billions of dollars in funding only to achieve the niche status that VM reached over 20 years of neglect.
There's certainly no marketing of IBM's killer apps by IBM. OfficeVision and CMS Infocenter are now almost dead, sad to say, but VM guest and client/server capabilities are not advertised in the media or marketed by the branches.
Outside the Endicott organization, there seems to be little realization that VM is IBM's best opportunity for mainframe-based web services. A few years ago, IBM asked me to address the MVS development organization in Poughkeepsie NY. I discussed their lovely full-page ads about S/390 based web servers. These were all artists mockups because they couldn't bring themselves to promote the mainframe web servers appearing in the real world - which happen to run on VM, not on MVS (sorry, I know I'm supposed to call it OS/390, but the habit is hard to break).
I want IBM to develop VM and its products in specific ways, however the product that's already there is excellent and deserves marketing support, plus a unequivocal statement from IBM senior management and branch marketing that it is vital and strategic.
I've watched a lot of IBM's market go to other vendors, and it's hard to blame anybody but IBM for their indifferent marketing and faint-heartedness. I exclude the IBM Endicott organization, which is doing the best it can against the mind-set rampant elsewheres in IBM.
I must emphasize, and IBM must understand, that customers will not migrate from VM to OS/2, AIX, AS/400, MVS, or Lotus (and therefore make it acceptable to IBM to let VM fade away). This is sheer hallucination. When I see users move an application off VM, they always move it to a non-IBM platform, and I'm certain this is true almost everywhere. IBM must wake up to the fact that they have to protect their VM product line because that revenue will otherwise go to Microsoft, Sun or some other competitor.
Some years ago, I walked Pat Hume, then manager of much of IBM's client-server product line, through Merrill Lynch offices at New York's World Financial Center to show her the non-IBM equipment appearing on most desks These desks all used to have 3270 terminals on them, connected to VM, and now have Suns. Not AIX. Not OS/2. (True, many of them were just being used as very expensive terminals running tn3270 into VM.)
IBM must abandon their illusions in this area very soon, before people manage to move the rest of their applications off VM and the IBM product line. These users are having difficulty, with some major flops, but they'll keep trying as long as it seems the most attractive option for them to pursue.
[ Prior page | Next page | Top page ]